You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Not able to reproduce semantic segmentation result on PASCAL VOC val with MMSegmentation.
I used the code from https://github.com/MenghaoGuo/-EANet/blob/main/model_torch.py and modified nothing despite replacing the backbone with the one in MMSegmentation. After training EANet and PSPNet with several sets of configs, the result is that the mIoU of EANet is always a little bit below PSPNet, e.g. 73 vs 75.
Any suggestions?
The text was updated successfully, but these errors were encountered:
Thanks for your attention.
We don't use mmsegmentation to train PASCAL VOC dataset. I suggest you can reproduce cityscapes or ADE20K results by using mmsegmentation.
If you want to reproduce PASCAL VOC result, i suggest you use https://github.com/XiaLiPKU/EMANet/ and replace it network.py, train.py, settings and eval.py by this repo.
If you want to test the model directly. You can download the pretrain model and use test.py to get test results.
Best.
Not able to reproduce semantic segmentation result on PASCAL VOC val with MMSegmentation.
I used the code from https://github.com/MenghaoGuo/-EANet/blob/main/model_torch.py and modified nothing despite replacing the backbone with the one in MMSegmentation. After training EANet and PSPNet with several sets of configs, the result is that the mIoU of EANet is always a little bit below PSPNet, e.g. 73 vs 75.
Any suggestions?
The text was updated successfully, but these errors were encountered: